In this paper we propose to use an aggregation rule derived from the Group Decision Theory, and based on the ranks of a set of individual indicators, for building a family of composite indicators for the security of energy supply. This family of composite indicators depends on a parameter associated with the risk-averse level of the decision maker, which can hence vary continuously from risk-prone to risk-averse. It represents a valuable and objective instrument to evaluate the degree of security of energy supply of different countries without entering into controversial matters related to the choice of the weights. This methodology can be used to aggregate either ranks or normalized values of the individual indicators. We apply it to a set of individual indicators, accounting for different dimensions of the security of supply and derived from the outputs of an energy system model. We study the evolution in time of the countries' performances, as well as the propagation of the uncertainties associated with the individual indicators to the composite. We also studied the robustness of such composite indicators with respect to the risk-averse level.